scientific correspondence devices accessible to everyone (M. D. S., unpublished data). These could replace the existing servo levitation devices for some applications.
a mN
mN
b mN
A. K. Geim*, M. D. Simon†, M. I. Boamfa*, L. O. Heflinger† *High Field Magnet Laboratory, University of Nijmegen, 6525 ED Nijmegen, The Netherlands e-mail:
[email protected] †Department of Physics and Astronomy, University of California at Los Angeles, Los Angeles, California 90095, USA 1. 2. 3. 4.
Earnshaw, S. Trans. Camb. Phil. Soc. 7, 97–112 (1842). Brandt, E. H. Science 243, 349–355 (1989). Berry, M. V. &. Geim, A. K. Eur. J. Phys. 18, 307–313 (1997). Thomson, W. (Lord Kelvin) Reprints of Papers on Electrostatics and Magnetism (Macmillan, London, 1872). 5. Braunbeck, W. Z. Phys. 112, 753–763 (1939). 6. Beaugnon, E. & Tournier, R. Nature 349, 470 (1991). 7. Geim, A. Phys. Today 51, 36–39 (September 1998). 8. Arkadiev, A. Nature 160, 330 (1947). 9. Ikezoe, Y. et al. Nature 393, 749–750 (1998). 10. Simon, M. D. et al. Am. J. Phys. 65, 286–292 (1997). 11. Boerdijk, A. H. Philips Tech. Rev. 18, 125–127 (1956).
Extraocular magnetic compass in newts Geomagnetic orientation is widespread among organisms, but the mechanism(s) of magnetoreception has not been identified convincingly in any animal1. In agreement with biophysical models proposing that the geomagnetic field interacts with photoreceptors2–4, changes in the wavelength of light have been shown to influence magnetic compass orientation in an amphibian, an insect and several species of birds (reviewed in ref. 5). We find that light-dependent magnetic orientation in the eastern red-spotted newt, Notophthalmus viridescens, is mediated by extraocular photoreceptors, probably located in the pineal complex or deeper in the brain (perhaps the hypothalamus). Experiments investigating shoreward magnetic compass orientation have demonstrated that the newt’s perception of the direction of the magnetic field is rotated 90o under long-wavelength (greater than 500 nm) light5,6. We recently trained newts under natural skylight to aim for the shore by placing them for 12–16 hours in waterfilled tanks with an artificial shore at one end5,7. The magnetic orientation of individual newts was then tested in a circular, visually symmetrical indoor arena under depolarized light. Under full-spectrum light (from a xenon arc source), they exhibited bimodal magnetic orientation parallel to the shoreward axis in the training tank (Fig. 1a,d). In contrast, under long-wavelength light, they orientated themselves perpendicular to the shoreward direction (Fig. 1b,e). To demonstrate that the 90o shift in orientation under long-wavelength light was 324
c
mShore d
mN"
mN"
mN"
mShore e
P < 0.001 g
mShore f
P < 0.001 mShore
h
mShore
P < 0.001 Figure 1 Effects of long-wavelength light and head caps on bimodal magnetic orientation in newts. a–c, Predicted orientation of newts (double-headed arrow) and their perception of the direction of the magnetic field (single-headed arrow)5,6. Training tanks have the shore towards magnetic north (mN); circular test arenas show the predicted response of the newts under either full-spectrum (beige) or long-wavelength light (yellow). a, Full-spectrum training and testing: newts should perceive the shore to be towards magnetic north and exhibit bimodal magnetic orientation along the shoreward axis. b, Full-spectrum training, long-wavelength testing: newts’ perception of magnetic north in testing, and their orientation in the test arena, should be rotated 90º (mN8) from magnetic north during training. c, Long-wavelength training and testing: newts’ perception of the magnetic field should be rotated 90º relative to the actual field during training and testing. Their perception of the magnetic field in the arena would be the same as in the outdoor tank. d–h, Results. Data points show the magnetic bearing of a newt tested in one of four symmetrical alignments of an Earth-strength magnetic field (magnetic north is geographic north (gN), east (gE), west (gW) or south (gS)). The magnetic field was altered by two orthogonally orientated, double-wrapped, Ruben’s coils around the test arena7. Data are plotted with respect to the magnetic direction of shore (mShore) in the training tank (shore direction, 360º). Double-headed arrows indicate mean axis of orientation with the mean axis length, r, proportional to the strength of orientation (diameter of the circle corresponding to r41). Dashed lines indicate 95% confidence intervals for the mean axis. Distributions are significant at P*0.05 or less by the Rayleigh test and P-values between circle plots indicate significant differences between distributions (Watson U 2 test). d, Newts trained under natural light and tested under full-spectrum light orientated along the shoreward axis5. e, Newts trained under natural light and tested under long-wavelength light orientated 90º from the shoreward direction (filled circles; tested under broadband long-wavelength (à500 nm) light; filled squares, tested under a 550-nm light, 40 nm bandwidth, 12.550.1 log Quanta cm12 s11; ref. 5). f, Newts trained and tested under long-wavelength light orientated along the shoreward axis5. g, After training under natural light, clear-capped newts tested under long-wavelength light orientated ~90º from the shoreward direction. h, After training under natural light, newts with long-wavelength-transmitting caps orientated along the shoreward axis under long-wavelength light.
due to a direct effect of light on the newts’ perception of the magnetic field, we trained newts under long-wavelength light by covering the training tank with a long-wavelength-transmitting gel filter (two layers of Lee #101)5. Under long-wavelength light, these newts orientated themselves parallel to the shoreward axis, indicating that they had learned the direction of the shore with respect to the rotated magnetic information under long-wavelength light (Fig. 1c,f). As well as ocular photoreceptors, newts have extraocular photoreceptors in the pineal complex8 and possibly the hypothalamus9. To determine which photoreceptors are involved in the magnetic compass response, we manipulated the wavelength of light reaching the extraocular photorecep© 1999 Macmillan Magazines Ltd
tors. Small round ‘caps’ (5 mm in diameter) were attached to the dorsal surface of the head of each newt using cyanoacrylate glue, and remained in place during both training and testing. Equal numbers of newts were capped with either a clear filter (Lee #130) or a filter that transmitted only long-wavelength light (equivalent to two layers of Lee #101). The caps were positioned to alter the spectral properties of light reaching the pineal and surrounding structures, whereas light reaching the eyes was unaffected. Clear-capped newts were tested to control for any nonspecific effects of the caps on the newts’ orientation behaviour. All newts were trained outdoors under natural skylight and tested for magnetic orientation in the testing arena under longNATURE | VOL 400 | 22 JULY 1999 | www.nature.com
scientific correspondence Table 1 Magnetic orientation of newts in the ‘cap’ experiments U2
r
n
PR
172–352 77–257
0.49 0.54
11 11
*0.1 *0.05
0.346
*0.001
clear yellow
62–242 162–342
0.76 0.61
10 12
*0.001 *0.01
0.472
*0.001
clear yellow
77–257 169–349
0.58 0.57
21 23
*0.001 *0.001
0.753
*0.001
Shore direction
Cap type
Magnetic axis (º)
West (shore, 270º)
clear yellow
North (shore, 360º) Pooled data (shore, 360º)
PW
Capped newts were trained to exhibit shoreward magnetic compass orientation in an outdoor tank with the shore direction to either the west or the north, and tested for magnetic orientation as in previous experiments5,7 under long-wavelength (à500 nm) light. Magnetic bearings were doubled before statistical analysis. For the pooled distributions (Fig. 1g,h), magnetic bearings from newts trained to the west shore were rotated by &90o so that the shoreward direction for both groups was at 360o. Magnetic axis is the mean for the bimodal distribution; r, mean vector length; n, sample size; PR, probability by the Rayleigh test; U 2, Watson U 2 test for differences between two distributions; PW, probability by the Watson U 2 test. For statistical analyses of the data shown in Fig.1d–f, see ref. 5.
wavelength light. Newts with clear caps were predicted to orientate along an axis perpendicular to the shoreward direction (like newts trained under full-spectrum light and tested under long-wavelength light; Fig. 1e). In contrast, the predicted orientation of newts with ‘long-wavelength’ caps depended on whether the caps altered the wavelength of light reaching the photoreceptors involved in the magnetic compass. If so, these photoreceptors would have been exposed to long-wavelength light in both training and testing, and the newts should have orientated along the shoreward axis (as in Fig. 1c, f). If not, the photoreceptors would have been exposed to full-spectrum light during training and long-wavelength light during testing. In this case, the newts should have orientated perpendicular to the shore (as in Fig. 1b,e). Newts with clear caps orientated perpendicular to the shoreward magnetic axis under long-wavelength light, indicating that the caps did not alter the orientation response (Table 1; Fig. 1g). In contrast, newts with long-wavelength-transmitting caps exhibited bimodal magnetic orientation parallel to the shoreward axis in the training tank (Table 1; Fig. 1h). Covering the dorsal surface of the newt’s head with a long-wavelength-transmitting filter therefore mimicked the effect of long-wavelength training on the newt’s shoreward compass response. Thus, extraocular photoreceptors are involved in the newt’s light-dependent magnetic compass. The effects of light on magnetic orientation in birds are different from those in newts (birds have been disorientated under wavelengths beyond 590 nm)5. Results from physiological studies indicate that both the avian visual system and the photosensitive pineal are sensitive to magnetic fields5, but behavioural experiments on birds from which the pineal has been removed suggest that it is not required for avian magnetic compass orientation10,11. It would not be surprising, therefore, if the photoreceptors underlying such orientation are in different locations in salamanders and birds (extraocular and ocular, respectively). Alternatively, our findings and the results of experiments on pinealectomized birds10,11 NATURE | VOL 400 | 22 JULY 1999 | www.nature.com
could both be explained by the involvement of deep brain photoreceptors in a lightdependent magnetic compass. M. E. Deutschlander*, S. C. Borland, J. B. Phillips The Center for the Integrative Study of Animal Behavior, Department of Biology, Indiana University, Bloomington, Indiana 47405, USA *Present address: Department of Biology, University of Victoria, P.O. Box 3020, Victoria, British Columbia V8W 3N5, Canada e-mail:
[email protected] 1. Wiltschko, R. & Wiltschko, W. Magnetic Orientation in Animals (Springer, Berlin, 1995). 2. Schulten, K. in Advances in Solid State Physics (ed. Treusch, J.) 61–83 (Viewig, Braunschweig, 1982). 3. Leask, M. J. M. Nature 267, 144–145 (1977). 4. Edmonds, D. T. Proc. R. Soc. Lond. B 263, 295–298 (1996). 5. Deutschlander, M. E., Phillips, J. B. & Borland, S. C. J. Exp. Biol. 202, 891–908 (1999). 6. Phillips, J. B. & Borland, S. C. Nature 359, 142–144 (1992). 7. Deutschlander, M. E., Phillips, J. B. & Borland, S. C. Copeia (in the press). 8. Adler, K. Photochem. Photobiol. 23, 275–298 (1976). 9. Foster, R. G., Grace, M. S., Provencio, I., DeGrip, W. J. & Garcia-Fernandez, J. M. Neurosci. Biobehav. Rev. 18, 541–546 (1994). 10. Schneider, T., Thalau, H.-P., Semm, P. & Wiltschko, W. J. Exp. Biol. 194, 255–262 (1994). 11. Maffei, L., Meschini, E. & Papi, F. Z. Tierpsychol. 62, 151–156 (1983).
Dating the origin of HIV-1 subtypes The moment in history when subtypes of the human immunodeficiency virus HIV-1 became distinguishable is hotly debated1,2. Zhu et al.3 have provided a unique HIV-1 sequence from 1959, known as ZR59. Based on the position of ZR59 in phylogenetic trees and its distance from the common node of the viral B/D/F-subtype lineages, they suggest that subtypes B and D have evolved from a single introduction “not long before 1959”, and that HIV-1 group M viruses probably shared a common ancestor “in the 1940s or the early 1950s”. Here we caution against such precise dating of these evolutionary events. Evolutionary rates among viruses are unlikely to be equal and constant over time, as the sequences presented by Zhu et al. © 1999 Macmillan Magazines Ltd
testify, with subtype B sequences from 1983–94 being significantly closer to the B/D/F node than are subtype D sequences sampled in the same period. We have retrieved 95 HIV-1 strains (75 B, 12 D and 8 F) from the GenBank datadase, for which our only selection criteria were that sequences covered the same genetic regions as ZR59abc and that sampling years were reported (1983–97). A linear regression analysis revealed that there was a significant (P*0.001) but weak (r 240.07) positive correlation between sampling years and sequence distances to the B/D/F node, with a slope (divergence rate) of 0.0009 or 0.0008 (including or excluding ZR59, respectively). Although a molecular clock is operational during HIV-1 evolution in an AIDS pandemic4, it is of limited value for dating the B/D/F node or the ancestor of group M viruses5: when we estimated the year the B/D/F lineages diverged, despite the 95% confidence interval of more than 100 years, we arrived at the year 1913 by assuming that each sequence was statistically independent and that HIV-1 evolution was not influenced by population bottlenecks. As there is no reliable way to control for the validity of these two assumptions and to narrow the confidence interval, such calculations can be dismissed as uninformative. When the limitations of the molecular clock are ignored and no confidence intervals for node dates are derived, all we have to work with is the genetic distance of the ZR59 sequence to the nodes. The mean sampling year of the contemporary sequences was 199055 (s.d.) and their mean distance to the B/D/F common node was 0.07150.017, which is only 2–3 times higher than we and Zhu et al.3 estimate for the ZR59 sequence. Even if the line of reasoning pursued by Zhu et al. is accepted, the time between the B/D/F node and 1959 should be equal to or half that passed between 1959 and 1990, leading to a date for the B/D/F node of either 1928 or 1944. For the group M node, a date decades before that is possible, but it is hard to date as its position depends on the phylogenetic method used. In our analysis, branching patterns of ZR59 and the position of the group M node (Fig. 1a) are similar to those of Zhu et al., but other methods yield a completely different position for the M node5. We uncovered several contemporary HIV-1 sequences that are equidistant or closer than the ZR59 sequence to the nodes of group M (Fig. 1b) and B/D/F lineages (Fig. 1c). Their existence makes dating of the nodes shortly before 1959 more cumbersome. Variation among HIV-1 sequences is increasing over time yet, in any year, sequences can be found that are close to the ancestor4. The Poisson model of the molecular clock implies variation in evolution rates among lineages4,6. Together with massive 325
